I imagine that some nonattendees see Apple's WWDC (Worldwide Developer Convention) as a huge pep rally where Steve and the rest of the Apple brass take turns preaching to the converted. That's just wrong; it isn't huge at all. Really, the WWDC crowd isn't there to get pumped up. Apple Developer Connection memberships and WWDC tickets are not bargains. But attendees pay to get in so they can get paid far more after they leave.
Each year when I make preparations for the WWDC, lessons from the prior year help me zero in on subject areas that are relevant to my readers. A solid half of WWDC focuses on Apple's accessible implementations of technical standards that are applicable across platforms. I recently realized that I come away from WWDC each year with core knowledge that's universally useful. I learned about VLANs at WWDC. One session allowed me to make sense of single sign-on. WWDC is where I learned Python. This year I'll learn Apple's new approach to running a mix of 32- and 64-bit applications. I hope to discover how the Mac's Core Microarchitecture CPU, Extensible Firmware Interface, and Trusted Platform Module are exposed to developers. And although this will certify that I'm certifiable, I am salivating at the prospect of seeing hands-on sessions with x86 assembly language. A fair portion of the knowledge I'm looking to gain isn't specific to the Macintosh or to OS X.
The applicability of the knowledge transferred at WWDC will be especially broad this year because Apple is set to turn a corner that I predicted: It is one step away from turning the Mac into the world's first universal x86 platform. Thanks to Parallels Desktop, an US$80 virtualization solution that will sell a million Macs all by itself, every Intel-based Mac will run every 32-bit x86 OS, from DOS to Longhorn Server, from FreeBSD to Suse Linux Enterprise System, at (truly) near-native speed and with no need to reboot to switch OSes. A cynic could say Apple is "cheating" its way to OS universality by blocking the use of OS X on non-Mac PCs. It's not cheating. Any first-tier system maker is free to create its own OS, lash that OS exclusively to boxes carrying its brand, and try to beat Apple at its own game. Until somebody does that, Apple owns bragging rights for selling the only box that will run all popular commercial and open source operating systems. You can't possibly imagine how fast Windows (or any x86 OS) runs as a guest of OS X Tiger.
I said that Apple is one step away from universality, and that step is a doozy: 64-bitness. This is a far tougher nut to crack for x86 than it was for PowerPC. The Mac server and workstation to come -- see The Enterprise Mac blog for detailed predictions -- will be 64-bit machines in brain and body, but running Tiger as it exists today, Intel Xserve and "Mac Pro" (or whatever they'll call Power Mac G5's replacement) will arrive handicapped, capable of taking advantage only of the new systems' greater memory capacity. That is a dreadful substitute for genuine 64-bitness -- do a site search with my name and "64" and you'll find out what I mean -- but it may be what emerges from WWDC and the state of the Mac art until Apple's Leopard OS goes public next year. For this WWDC, I'd be content to find that Apple has paved the way for its developers and high-end server and workstation users to build their own bridges over whatever gaps exist between OS X and new Intel Macs' hardware capabilities. I expect to be content.